Analyzing Expression Profiles from Single Stem Cells Using the Single Cell-to-Ct™kit

1 12/23/2011 | Life Technologies™ Proprietary and confidential

Analyzing Expression Profiles from

Single Stem Cells Using the Single Cell-

to-Ct™ kit

Ron Abruzzese, Ph.D.

Life Technologies

Austin, Texas

2 12/23/2011 | Life Technologies™ Proprietary and confidential| Life Technologies |

Why Study of Single Cells?

Cells are not homogenous

Research into rare cell

or event

• Circulating metastatic cells

• Fetal cell in maternal blood

• Event within a library

Research using

scarce, precious

sample

• Archival tissue (FFPE)

• Clinical sample (fresh tumor)

• Biomarker discovery

•Data from ensemble averaging in cell

measurements can be misleading

• Interesting Biological Research Applications

Learn more about the

Single Cell-CT Kit (400 Reactions)

http://bit.ly/stem-cell-road-show



https://products.invitrogen.com/ivgn/product/4458236?ICID=search-product


Develop a complete workflow to obtain

statistically relevant single cell data

Dynabeads®

CD3/CD28

Naive or

resting

T cell

Benefits

•Optimized reagents

•Superscript Vilo

•Reformulated pre-Amp

•Small volumes & no

dilution enable use of

entire lysate sample in

subsequent RT/PreAmp

rxns







•500 L Single Cell Lysis Solution (store at 4ºC)

•50 L Single Cell Stop Solution (store at -20ºC)

•50 L Single Cell DNase I (store at -20ºC)

•150 L Single Cell VILO™ RT Mix (store at -20ºC)

•75 L Single Cell SuperScript® RT (store at -20ºC)

•265 L Single Cell PreAmp Mix (store at -20ºC)

Kit Components 50 and 400 reaction kits







Starting Material

Samples can be obtained through

− FACS

− Dilution

− Laser capture microdisection (not tested internally)

− Physical selection (eg bead based)

− Laser Ablation (Cyntellect Leap System; not tested internally)

− Mouth pipetting

Up to 10 cells can be used

Input volume of cell (s) should be less than 1 ul

We have validated 5 cell lines (including hESC) for this kit and >20 for the parent kit







Product Performance

12

14

16

18

20

22

24

26

28

CT

1 Cell

Equivalents

Single Cells 100 cells

20.4

21.6

13.6

Single Cell Detection Occurs with expected sensitivity (6.6 Cts difference from 100

cells). Technical reproducibility of the 100 cell samples and single cell equivalents

is tight (CV of Cell Equivalents is small). Variability of single cells is due to

biological variability in single cells

N=84 single cells






7 12/23/2011 | Life Technologies™ Proprietary and confidential| Life Technologies Proprietary & Confidential | 12/23/2011

SV25 (Olig2-EGFP), a derivative of BG01• Platform line maintains expression of hESC markers







Grow ESCs on Geltrex plates in CM

until 80-90% confluent

Change media to SFM for 24 hours.

Culture in NAA media until

confluent, changing daily.

Split cells 1:2 using TrypLE, culture

on Geltrex plates.

Once confluent use collagenase,

create neurospheres 150 – 250 um

in Ultra Low Attachment plates.

When cells start to attach to the

plate, split 1:2 onto Geltrex plates.

Continue culturing on Geltrex

plates, splitting 1:2 every 2-3 days.

Dissociate rosettes into single

cells

Day 0 ESC

Day 1 ESC

Day 2 Differentiating ESC

Day 6 Neurospheres

Day 13 Neurospheres

Day 17 Attached

Neurospheres

Day 20 Rosette formation

Day 22 NSC derived

ESC to NSC Workflow 1575 m

Day 0 Day 3

Day 7 Day 11

Day 17 Day 18

Day 21 Day 21

157 mDay 24 GFP Overlay

| Life Technologies Proprietary & Confidential | 12/23/2011

Learn more about the Single Cell-CT Kit (400 Reactions)





Tested Single Cell Analysis Workflow

• Look at single cells to more closely define profiles, use cells to cell isolation

in 96-well plates, to qPCR in 384-well plates

0 cells 100 cells

10 genes/cell10 plates/time point

•30 “0” cell samples



Cells-to-CT™ TaqMan®

PreAmp MMx

Gene

Expression

Master Mix

VILO RT

Superscript®


Embryonic Stem Cells







Single cell analysis or Embryonic Stem Cells

100 cells (average CT): 13.7 + 0.2

1 cell low cluster (36 cells): 19.2 + 1.3

1 cell high cluster (48 cells): 27.4 + 1.0

Average CT (84 cells): 21.4 + 4.2

Single cell equivalents (100 samples):

22.8 + 0.3

• Analyzing gene expression profiles en masse gives an average profile

• Obscures or potentially obliterates any differences in single cells

100 cells1 cellACTB








Single cell analysis or Embryonic Stem Cells• Gene variability - large expression range for one gene; size variations do

not account for this, but cell cycle dependent regulation may

100 cells1 cell

ACTB

100 cells1 cell

OCT4

• Cell-to-cell variability - expression profiles are not the same in every cell

• See small sub-populations (OCT4 low expressers)

“technical” variability

22.8±0.3 (6.2 from 100

cell samples)

• Technical variability (from method of detection) needs to be identified

(low here)








10

20

30

40

CT

10

20

30

40

CT

CT

Day 0

Day 14

Day 24

Variation in expression level in single cells

10

20

30

40

GA

PD

GF

AP

NE

S

PA

X6

PO

U5F

1 T

TU

BB

3

UT

F1

ZF

P42







Noise - Effects on Normalization

• Noise is an inherent part of a biological system and results in cell-to-cell differences

• Extrinsic noise - variations of the levels of transcription factors, polymerases etc. -

results in cell to cell differences for total fluorescence (or total levels of

transcription)

• Intrinsic noise - variation introduced from the act of transcription itself - results in

differences in the levels of independently expressed fluorescent proteins that are

under identical promoters

• Noise causes differences that calls into question the use of “normalization”

• When analyzing a population noise is averaged out making “normalization” more

appropriate.

Elowitz et al., Science 2002| Life Technologies Proprietary & Confidential | 12/23/2011







100 Cell Data From Day 0

5

10

15

20

25

30

35

GAPDH NES POU5F1 TUBB3 ZFP42

CT

Thirty 100 cell samples show similar expression levels as demonstrated by small center quantiles (left).

Normalized expression levels of each gene to GAPDH expression levels remove some of the sample to

sample variability as shown by smaller box and whisker (right) and show that the gene “profiles” of each

sample are very similar.

•100 cell samples have similar expression levels which

tighten when normalized

-10

-5

0

5

10

15

20

CT

NES-

GAPDH

POU5F1-

GAPDH

TUBB3-

GAPDH

ZFP42-

GAPDH







Single Cell Data From Day 0

•Single cell samples give wide range of expression levels which spreads

out further when normalized

900 single cell samples show a wide range of expression levels shown by the large box and

whiskers (left). After normalization (right), box and whisker sizes increase as does the

number of outliers

15

20

25

30

35

40

45

GAPDH NES POU5F1 TUBB3 ZFP42

CT

-10

-5

0

5

10

15

20

NES-

GAPDH

POU5F1-

GAPDH

TUBB3-

GAPDH

ZFP42-

GAPDH

CT







Conclusions and Impacts• There are significant differences from cell-to-cell

• Analyzing gene expression en masse gives an average profile

and masks differences and variability (gene-to-gene and cell-to-

cell)

• Small populations are lost when large populations are averaged

• The TaqMan Single Cell-to-Ct kit:


• Optimized reagents provide a simplified workflow for expression

analysis of single cells by qRT-PCR

• Enables transfer of entire cell into each step

• No sample is lost during the reaction which occurs in a single

tube

• Enables the acquisition of statistically significant data sets







Normalization Conclusions

•When analyzed en masse, variation in expression level is

reduced when results are normalized to reference genes.

•Expression levels of each gene vary independently within

a single cell

•In single cells normalization increases the variation in

calculated expression level.

•These normalized values are not the same within each

cell and vary depending on the genes compared.

•These results suggest that normalizing single cell data is

not an accurate method of analysis.







Acknowledgements

Ron Abruzzese

Richard Fekete

Laura Chapman

Dan Kephart

Andrew Lemire

Penn Whitley

Elena Grigorenko

Ying Liu

Chad MacArthur

Gothami Padmabandu

Jon Chesnut

Mahendra Rao

Janice Au-Young

David Keys

Jonathan Wang

Life Technologies:

2130 Woodward St.,

Austin, TX

12 Gill St. Suite 4000

Woburn, MA

25791 Van Allen Way,

Carlsbad, CA

850 Lincoln Center Dr.,

Foster City, CA


Life Technologies, Applied Biosystems and Ambion products are for Research Use Only. Not for

use in diagnostic procedures.

The trademarks mentioned herein are the property of Life Technologies Corporation or their

respective owners. TaqMan is a registered trademark of Roche Molecular Systems, Inc.

AmpliGrid is a trademark of Beckman Coulter Inc.

NOTICE TO PURCHASER: Limited Use Label License

The products shown in this presentation may be covered by one or more Limited Use Label

License(s). Please refer to the respective product documentation or the Applied Biosystems

website under www.appliedbiosystems.com for the comprehensive license information. By use of

these products, the purchaser accepts the terms and conditions of all applicable Limited Use

Label Licenses. These products are sold for research use only, and are not intended for human or

animal diagnostic or therapeutic uses unless otherwise specifically indicated in the applicable

product documentation or the respective Limited Use Label License(s).

© 2010 Life Technologies Corporation. All rights reserved.

Legal StatementsLearn more about the






www.lifetechnologies.com



http://www.lifetechnologies.com/us/en/home.html




Analyzing Expression Profiles from Single Stem Cells Using the Single Cell-to-Ct™kit

Technology

cell cells

single cell dnase

cell samples

rare cell fetal cell

cell lines

thesingle cellct kit

single cell preamp mix

variability of single